Transformer load prediction method

• Main Authors: AI HONGKE, PENG XIAOLEI, TAN HUIZHENG, MIAO QIPENG, GAO LIMING, MA YAYUN, YAN RUITAO, JIANG XIAOCHAO, DU PENG, XUE SHUAI, CHEN ZHIYU, TIAN XINCHENG, YAN KUN, LU ZEHAN, CHENG DECAI, WANG DONGRUI, GAN JINGFU
• Format: Patent
• Language: Chinese; English
• Published: 13.09.2022
• Subjects: CALCULATING; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTINGELECTRIC POWER; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRICITY; GENERATION; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS; SYSTEMS FOR STORING ELECTRIC ENERGY; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention relates to a transformer load prediction method, and belongs to the technical field of power load prediction methods. According to the technical scheme, a transformer historical daily load curve is used as an original sample, an improved clustering algorithm is used for recognizing and removing abnormal load curves in the original sample, several types of typical daily load curve samples are obtained, an approximate number of samples are taken from each type of screened samples to form a neural network training set, and the neural network training set is used as a neural network training set; and training the neural network and establishing an accurate prediction model of the daily load curve of the transformer. The method has the advantages that interference of abnormal data samples on the neural network is eliminated, screening and classification of available samples are achieved at the same time, the training sample scale is reduced, meanwhile, the category with the small number of samples is